Re: {Fixed] Unexpected scrolling
Jack <[email protected]>
| Newsgroups | gmane.comp.gnome.apps.balsa |
|---|---|
| Message-ID | <[email protected]> |
On 2018.04.12 00:33, Peter Bloomfield wrote: > On 04/08/2018 06:23:47 PM Sun, Peter Bloomfield wrote: >> An irritating bug has been mentioned on the list, e.g. by Jack[0] >>> This last one is different from the issue that the first time I >>> click anywhere in the body of a message (default layout, lower >>> right pane) it seems to scroll the message down just enough to move >>> the header area out of view. This happens whether I am displaying >>> a plain text or an html message. >> Just hoping that it would go away was tempting, but hasn't solved >> the problem :-) I finally found it aggravating enough to track it >> down, and file a bug[1] against Balsa describing what I found. I'm >> testing a workaround, and will post a patch to the bug if it seems >> solid. >> >> All contributions welcome! >> >> Peter >> >> [0] >> https://mail.gnome.org/archives/balsa-list/2015-April/msg00000.html >> [1] https://bugzilla.gnome.org/show_bug.cgi?id=795085 > > Embarrassingly simple fix! Long, long ago, we had to wrap a message > in a GtkViewport before adding it to the GtkScrolledWindow that makes > it, well, scrollable. Then GtkScrolledWindow added a 'convenience' > feature, adding the GtkViewport automatically when needed, so we > stopped doing it for ourselves. Unmentioned in the documentation of > the feature is that it adds another 'convenience', namely scrolling a > child widget to the top of the scrolled window when it gets the > focus, for example by--you guessed it--clicking on it. So thank you, > GtkScrolledWindow, for all those years of 'convenient' scrolling! > > A fix in master reverts that old change; we go back to wrapping the > message in a viewport, and clicking no longer leads to unwanted > scrolling. > > Happy mailing (with clicks), > > Peter Well, I've finally managed to beat "real life" into a modicum of submission and managed enough time to compile a new version of Balsa. Hooray! At least this fix works for me - I have another application where the bugs I experience seem to change every time I recompile, even from exactly the same source. I've got gremlins in my system, but at least they are staying clear of Balsa for now. :-) Thanks for following up on this. Jack